Hi Jordan,
let me have a stab at that one:
> 129 pediatric brain scans were used and consisted of children ages 8-12.
Sounds good ;)
> 1. run estimate and write using coxxxx files
> -Tissue probability Map: TPM.nii
> -Affine regularisation: average sized template
Did you generate the TPM.nii with TOM8, based on the NIH sample? I would
suggest you do that as you then do not use adult reference information
at all.
> Extended options
> -spatial normalization: High-dimensional:Dartel
> Dartel Template: Template_1_IXI550_MNI152.nii
I would not do that as you do not want to DARTEL your images at this
step anyway (and if, you would use adult reference information as you
don't have a pediatric DARTEL template). Leave that at low-dimensional.
> writing options (used same for GM,WM, and CSF)
> -native space:none
> -normalized:none
> Modulated normalized: non-linear only
> DARTEL export: affine
The only one you need is the last one, as this will write out your
images in an affine-registered way which you can then enter into later
TOM analyses. There is some debate as to whether you should use affine
registration only at this point, so there is no way to be absolutely
sure. In your specific setting, you could argue that you (if you use a
TOM8 TPM.nii) already have pediatric reference data so you might as well
use non-linear normalization, but that is something you will have to
decide (you would then set the second option to yes). I tend to stick
with affine. In any case, you do not want to use modulated data.
> I then left the rest at default
Sounds good.
> 2. estimate regression parameters: selected all GM, WM, CSF, and T1 images that were produced using the VBM8 estimate and write and input them to their appropriate branch.
As you realized, you will not get a complete TPM with all the 6
different background classes if you only enter GM, WM, CSF. I have not
checked new segment, but if you check cg_vbm8_run (around line 75) and
cg_vbm8_write (around lines 447) you will see that you can make VBM8
write out all tissue classes, which you can then again feed to TOM8.
> - input ages and gender in accordance to their input in the previous branches (GM,WM,etc)
... according with the order in which you entered the tissue classes, I
guess? You will also have to choose the order of age regression, for
which there is an appropriate help text.
> 3.Create new template: Selected the created TOM.mat file
> -selecte template creation method: average approach
With a larger group, probably the better option.
> - input ages and gender in accordance to their input in the previous branches (GM,WM,etc)
Yep.
> This then gave me 3 files labeled T1_Template_9.545.nii.
If you are getting three T1-templates, something went wrong (and this is
relevant as the T1 is scaled prior to averaging and thus treated
differently from the other tissue classes). You should get a T1-template
and tissue priors for the classes you entered.
> I apologize for the lengthy post, but I want to be absolutely certain this was performed correctly before I proceed further.
Wanting to be sure is always sensible, but there are rarely guarantees
for the "absolutely certain" level (including this mail :)
Hope this helps,
Marko
--
____________________________________________________
PD Dr. med. Marko Wilke
Facharzt für Kinder- und Jugendmedizin
Leiter, Experimentelle Pädiatrische Neurobildgebung
Universitäts-Kinderklinik
Abt. III (Neuropädiatrie)
Marko Wilke, MD, PhD
Pediatrician
Head, Experimental Pediatric Neuroimaging
University Children's Hospital
Dept. III (Pediatric Neurology)
Hoppe-Seyler-Str. 1
D - 72076 Tübingen, Germany
Tel. +49 7071 29-83416
Fax +49 7071 29-5473
[log in to unmask]
http://www.medizin.uni-tuebingen.de/kinder/epn/
____________________________________________________
|